GEOPHYSICAL RESEARCH LETTERS, VOL. 39, L22602, doi:10.1029/2012GL053624, 2012
Ocean temperature response to idealized Gleissberg and de
Vries solar cycles in a comprehensive climate model
Anne Seidenglanz,1 Matthias Prange,1 Vidya Varma,1 and Michael Schulz1
Received 21 August 2012; revised 15 October 2012; accepted 17 October 2012; published 17 November 2012.
[1] The 90-year Gleissberg and 200-year de Vries
cycles have been identified as two distinctive quasi-periodic
components of Holocene solar activity. Evidence exists for
the impact of such multi-decadal to centennial-scale variability in total solar irradiance (TSI) on climate, but
concerning the ocean, this evidence is mainly restricted to
the surface response. Here we use a comprehensive global
climate model to study the impact of idealized solar forcing,
representing the Gleissberg and de Vries cycles, on global
ocean potential temperature at different depth levels, after a
recent proxy record indicates a signal of TSI anomalies in
the northeastern Atlantic at mid-depth. Potential impacts of
TSI anomalies on deeper oceanic levels are climatically
relevant due to their possible effect on ocean circulation by
altering water mass characteristics. Simulated solar anomalies are shown to penetrate the ocean down to at least deepwater levels. Despite the fact that the two forcing periods
differ only by a factor of 2, the spatial pattern of response
is significantly distinctive between the experiments, suggesting different mechanisms for solar signal propagation.
These are related to advection by North Atlantic Deep Water
flow (200-year forcing), and barotropic adjustment in the
South Atlantic in response to a latitudinal shift of the westerly wind belt (90-year forcing). Citation: Seidenglanz, A.,
M. Prange, V. Varma, and M. Schulz (2012), Ocean temperature
response to idealized Gleissberg and de Vries solar cycles in a comprehensive climate model, Geophys. Res. Lett., 39, L22602,
doi:10.1029/2012GL053624.
1. Introduction
[2] Variations in total solar irradiance (TSI) on multidecadal to centennial time scales may constitute an important
contribution to climate variability on a societal time scale and
have been widely discussed [e.g., Gray et al., 2010].
Reconstructions of TSI based on cosmogenic radionuclides
have identified the Gleissberg (90 years) and de Vries
(200 years) cycles as two distinctive quasi-periodic and
non-stationary components in Holocene solar activity
[Knudsen et al., 2009; Ma, 2009]. Evidence of the surface
climatic and oceanic response to solar impact on these time
scales has been inferred from variations in sea surface temperature [e.g., Bond et al., 2001; Jiang et al., 2005], precipitation [e.g., Wang et al., 2005], and variations in sea level
1
MARUM–Center for Marine Environmental Sciences and Department
of Geosciences, University of Bremen, Bremen Germany.
Corresponding author: A. Seidenglanz, MARUM–Center for Marine
Environmental Sciences, University of Bremen, Leobener Straße, D-28359
Bremen Germany. ([email protected])
©2012. American Geophysical Union. All Rights Reserved.
0094-8276/12/2012GL053624
pressure and associated wind fields [Shindell et al., 2001;
Varma et al., 2011]. However, indications for the influence
of solar anomalies on deeper oceanic levels is scarce.
[3] A correlation between foraminiferal Mg/Ca and TSI
over the past 1200 years in the eastern North Atlantic is
indicative for a propagation of solar-induced temperature
anomalies also to sub-surface depths (900 m) [Morley et al.,
2011]. According to this study, temperature anomalies may
propagate from the subpolar into the subtropical gyre in
response to changes in solar-induced variations in surface
wind forcing. Furthermore, numerical experiments with
idealized solar forcing have been carried out, revealing
sensitivity of the Atlantic Meridional Overturning Circulation (AMOC) to TSI changes on the multi-decadal to centennial time scale [Park and Latif, 2012].
[4] To address the oceanic response to TSI variations, we
use a comprehensive global climate model. Sensitivity
experiments with idealized solar forcing, representing the
Gleissberg and de Vries solar cycles, are used to study the
impact of multi-decadal to centennial-scale solar variability
on ocean potential temperature at different depth levels.
Spatial variability in the temperature response to this idealized forcing is interpreted in terms of propagation pathways
of solar anomalies within the ocean. Contrary to previous
studies, which aimed at simulating the true temperature
response to the full range of solar variability during parts of
the Holocene [e.g., Cubasch et al., 1997; Weber et al., 2004;
Ammann et al., 2007], this study focuses on the mechanisms
of ocean response to idealized TSI at distinctive periods
which have been detected in paleoclimatic records. Specifically, we address the following questions: (1) Is there a
significant response in ocean temperature at the imposed
forcing frequencies? (2) Can solar-induced changes in temperature be tracked to sub-surface depth in accordance with
the proxy results? (3) What are the pathways of solar signal
transfer into the oceanic interior?
2. Experimental Design
[5] Sensitivity experiments with idealized solar forcing
were carried out using the NCAR Community Climate
System Model Version 3 (CCSM3), a comprehensive global
climate model consisting of ocean, atmosphere, land and
sea-ice components [Collins et al., 2006]. The CCSM3 has
successfully been employed in earlier studies regarding the
climate response to the 11-year solar cycle [Meehl et al.,
2008]. Considering the long integration time, the low resolution version with T31 spectral truncation [Yeager et al.,
2006] has been used. It corresponds to an atmospheric grid
spacing of 3.75 and 26 levels in the vertical. The ocean
model (POP) grid is resolved at 3.6 with a low-latitudinal
refinement in resolution (maximum of 0.9 at the equator)
L22602
1 of 6
L22602
SEIDENGLANZ ET AL.: OCEAN RESPONSE TO SOLAR FORCING
L22602
Figure 1. Coherency between ocean potential temperature (annual means) and TSI at the forcing periods of (a) 90 years
(representing the Gleissberg cycle) and (b) 200 years (representing the de Vries cycle), for three depth levels. Hatching indicates significant values at the 95% level. Critical values of coherency are 0.34 and 0.59 in the 90-year and in the 200-year
forcing experiment, respectively.
and 25 levels. The stratosphere in the CCSM3 has only a low
vertical resolution and does not include potential feedbacks
associated with changes in stratospheric ozone due to solar
UV (ultraviolet) variability [cf. Krivova et al., 2011]. Solar
forcing is implemented through a wavelength-independent
change in TSI, hence mostly affects the climate system
through shortwave absorption by the surface.
[6] Two sensitivity experiments with idealized TSI, each
integrated for 1000 years, were carried out. Solar forcing
varies sinusoidally with an amplitude of 1 Wm 2 (between
1364 Wm 2 and 1366 Wm 2) at the periods of the Gleissberg
cycle (90 years) and the de Vries cycle (200 years). This
forcing amplitude corresponds to a maximum estimate of
solar variations at multi-decadal to centennial time scales
[e.g., Steinhilber et al., 2009] and has been chosen to enhance
the detectability in the model experiments. We note, however, that substantial uncertainty regarding the magnitude of
past centennial-scale TSI variability exists [e.g., Lockwood,
2011]. To be comparable to each other, the two experiments were initialized from a quasi-equilibrated experiment
with late Holocene (2.5 ka BP) orbital forcing, as the integration of the 200-year solar forcing experiment is a continuation of a previously performed 700-year long simulation
with late Holocene boundary conditions [Varma et al., 2011].
Pre-industrial conditions were applied for the atmospheric
composition [Braconnot et al., 2007].
[7] The impact of TSI anomalies on the global oceanic
temperature field is studied by means of cross-spectral
analysis. (Squared) coherency between solar forcing and
annual means of temperature time series is determined at the
two forcing periods (90 and 200 years), and at specified
depth levels, in order to identify regions of significant
impact of TSI anomalies on ocean temperature. This is
extended by studying the lag of the temperature response to
TSI forcing at the two forcing periods to conclude about
possible propagation pathways of solar anomalies within the
ocean. The phase angle (f) is defined in the interval [ 180 ;
+180 ] due to the periodic behavior of the arcus-tangent
function. The phase angle is converted into a corresponding
phase lag (t) using t = f / 360 T, where T denotes the
2 of 6
L22602
SEIDENGLANZ ET AL.: OCEAN RESPONSE TO SOLAR FORCING
L22602
Figure 2. Phase angle (degrees) and lag (years) of ocean potential temperature to idealized TSI forcing at the surface for
(a) the 90-year forcing period, and (b) the 200-year forcing period. We assume that at each point changes in TSI lead the
variations in temperature (see Section 2). Hatching indicates significant coherency at the 95% level (cf. Figure 1).
forcing period in each experiment. Negative phase angles,
which result from the wrapping of f > 180 into the interval
[ 180 ; +180 ], are corrected by making the physically
plausible assumption that at each point changes in TSI lead
temperature variations. Cross-spectral analysis is based on
smoothed periodograms using a Daniell filter. Initially, time
series were tapered, linearly detrended, and had the mean
subtracted using the corresponding routines of the NCAR
Command Language (NCL) package, version 6.0.0 of
Brown et al. [2012].
3. Results
[8] The impact of TSI anomalies at the 90-year and 200-year
forcing periods on ocean temperature is investigated at the
surface, at mid-depth (1112 m) and at deep-water (1884 m)
level. For the Atlantic Ocean, the mid-depth level allows
for a qualitative comparison with the reconstructed temperature response of Morley et al. [2011], while the depth
of 1884 m represents North Atlantic Deep Water (NADW) in
the model.
[9] Coherency between temperature and solar forcing
(Figure 1) yields the spatially most homogeneous response at
the surface for both forcing periods, in tropical and subtropical regions, but also at higher latitudes in the Arctic Ocean
(90-year forcing) and in the North Atlantic region (200-year
forcing). With increasing depth, temperature sensitivity to
solar forcing reduces and a pattern of different spatial
temperature sensitivity between the two experiments emerges. The 90-year forcing induces a response in the South
Atlantic and Southern Indian and Pacific Ocean (1112 m
and 1884 m). In contrast, the 200-year experiment exhibits
3 of 6
L22602
SEIDENGLANZ ET AL.: OCEAN RESPONSE TO SOLAR FORCING
L22602
a significant temperature response in the North Atlantic
and subtropical South Pacific at mid-depth (1112 m), and
nearly the entire Atlantic and Arctic Ocean at the deep
level (1884 m).
4. Discussion
Figure 3. Phase angle (degrees) and lag (years) of ocean
potential temperature to idealized TSI forcing at 1884 m
depth for (a) the 90-year forcing period, and (b) the 200-year
forcing period. We assume that at each point changes in TSI
lead the variations in temperature (see Section 2). Hatching
indicates significant coherency at the 95% confidence level
(cf. Figure 1).
[10] Despite the fact that the two forcing periods differ
only by a factor of approximately two, the response patterns
in the oceanic interior differ significantly between the
experiments (Figure 1). To gain further insight into this
unexpected result, we turn to the phase lag of the temperature response to TSI forcing in order to deduce mechanisms
of signal propagation in the oceanic interior. We limit the
interpretation to regions for which the coherency between
TSI and temperature is significant at the 95% confidence
level.
[11] The surface-temperature response to the idealized TSI
forcing (Figure 2) occurs within 20 years and 44 years for
the 90-year and 200-year forcing, respectively. Moreover,
the two experiments reveal a positive correlation between
sea-surface temperature and TSI (not shown). Together with
the small lag seen in the temperature response, this is consistent with a direct, thermal forcing of the sea surface in
accordance with previous findings [e.g., Swingedouw et al.,
2011].
[12] At depth, the temperature lag to TSI forcing is compared between the two experiments at 1884 m (Figure 3).
Given the pronounced response especially in the Atlantic
Ocean at this depth, we focus on this region for the lag
analysis.
[13] In the 90-year forcing experiment, the temperature
response is visible as a zonal band at approximately 40 –
50 S in the South Atlantic (cf. Figure 1). The lag associated
with this response is short (f < 80 or t < 20 yrs; Figure 3a).
Varma et al. [2011] noted a latitudinal shift in Southern
Hemisphere westerly winds in this latitude band in
response to solar irradiance variations based on sensitivity
experiments with the same climate model and resolution.
Accordingly, centennial-scale intervals of higher solar
activity were associated with a poleward shift in the westerly
winds, and vice versa during solar minima. Our experiment
reveals a small positive correlation of potential temperature
with TSI anomalies (slope 0.1 C/(Wm 2); not shown) in
this region at the deep-water level as well as at the surface. This implies a wind-induced latitudinal shift of the
oceanic frontal system at these latitudes, involving a further southward penetration of North Atlantic Deep Water
(NADW) during solar maxima, thereby displacing colder
Circumpolar Deep Water southward in response to a barotropic adjustment to this shift in wind forcing. A similar
feature is visible in the 200-year forcing experiment, albeit
not statistically significant (Figure 1b). Neither of the two
experiments supports the solar-induced temperature variations at mid-depth in the northeast Atlantic as reconstructed by Morley et al. [2011]. We note, however,
elevated but statistically insignificant coherency values
along the northwest African margin from where the proxy
information stems.
[14] The 200-year forcing experiment reveals a gradually
increasing lag at 1884 m depth from the northeast Atlantic
(t < 30 years) towards the South Atlantic along the western
boundary (t 125 years; Figure 3b), as well as towards the
4 of 6
L22602
SEIDENGLANZ ET AL.: OCEAN RESPONSE TO SOLAR FORCING
L22602
Figure 4. Coherency between the streamfunction of the Atlantic meridional overturning circulation (AMOC) and idealized
solar forcing at the 200-year forcing period. Hatching indicates significant coherency at the 95% level. Black contour lines
show the long-term mean overturning streamfunction (in units of 106 m3 s 1).
Arctic Ocean (not shown). Previously performed sensitivity
experiments with idealized TSI forcing on multi-decadal
time scales revealed a sensitivity of the AMOC strength to
solar forcing, albeit with a forcing amplitude twice as large
as in our experiments (2 Wm 2) [Park and Latif, 2012].
Given the spatial pattern of response in the 200-year forcing experiment at 1884 m depth, we investigated the
coherency of the AMOC with idealized TSI in this experiment (Figure 4). Significant coherency can be detected at
subpolar northern latitudes (ca. 60 –80 N), reaching a
depth of 3600 m. However, the southward component of
the meridional overturning stream function at depth does
not show any significant response to the TSI forcing. The
pattern of increasing temperature lag at the deep level
(1884 m) nevertheless indicates a propagation of solar
anomalies with the deep western boundary current, taking
into account that the clock-wise circulation pattern in the
deep North Atlantic (cf. Figure 3b) is due to a shortcoming
in the model simulation of the deep western boundary
current in this region [cf. Prange, 2008]. Thus, the solar
signal becomes entrained into the NADW by deep convection and is subsequently propagated towards the South
Atlantic via the deep western boundary current without
involving any changes in the strength of the meridional
overturning circulation. Given the discrepancy with the
sensitivity seen in the AMOC strength when forced with a
twice-as-large TSI amplitude [Park and Latif, 2012], we
note that our experiments results may yield a qualitatively
different response pattern when excited by a different
forcing amplitude. However, investigating this aspect is
beyond the scope of our study.
[15] Recent simulations of the surface-climate response to
the 11-year solar cycle [Meehl et al., 2009; Bal et al., 2011],
including detailed representations of the stratosphere and the
stratospheric ozone response to ultraviolet irradiance variations, have resulted in a better match with observed climate
anomalies as compared to the sole dependence on TSI
forcing. Despite the fact that the stratosphere is only poorly
represented in the present experiments, the obtained results
are nevertheless promising as a similar model set-up (i.e.,
TSI forcing of CCSM3) was largely successful in simulating
the surface-climate response to solar variability, albeit with
lower amplitude [Meehl et al., 2008].
5. Conclusions
[16] Our experiments with idealized TSI forcing suggest a
significant sensitivity of ocean temperature down to a depth
of at least 1884 m at both forcing periods. Especially at
depth, the spatial expression of the temperature response
differs between the two experiments. While the 90-year
forcing induces a significant temperature response in the
mid-depth (1112 m) and deep (1884 m) South Atlantic
(40 –50 S), the 200-year forcing reveals the highest sensitivity to solar-irradiance changes in the northeast Atlantic
at mid-depth (1112 m), and nearly the entire Atlantic and
Arctic Ocean at the deep-water level (1884 m). Accordingly,
different mechanisms of solar signal propagation are proposed for the two experiments, which involve barotropic
adjustment in response to shifting westerlies in the South
Atlantic when forced with the 90-year period in TSI, and
southward advection by NADW in the 200-year-forcing
scenario. The previously detected solar-induced temperature
anomalies in the northeast Atlantic at the multi-decadal time
scale during the late Holocene derived from a proxy-based
reconstruction could not unambiguously be confirmed by
our experiments. Furthermore, future work is required to
more accurately estimate the climatic response to solar variability using models with a properly resolved stratosphere,
as has recently been suggested to be important on the 11year time scale. Our sensitivity analysis could not confirm
the sensitivity in AMOC strength to TSI forcing in previous
experiments with idealized solar forcing using a forcing
amplitude twice as high as the present simulations. Consequently, our findings illustrate the importance to systematically evaluate the role of the forcing amplitude in numerical
experiments in triggering different mechanisms of solar
signal propagation in the ocean.
5 of 6
L22602
SEIDENGLANZ ET AL.: OCEAN RESPONSE TO SOLAR FORCING
[17] Acknowledgments. We thank Dennis Shea (NCAR, Boulder,
Colorado) for providing details on the cross-spectral analysis procedure of
the NCL routines, as well as Ute Merkel (MARUM) for stimulating discussions on the topic. The CCSM3 model experiments were run on the
SGI Altix Supercomputer of the “Norddeutscher Verbund für Hochund Höchstleistungsrechnen” (HLRN). We are grateful to the two anonymous referees for their very constructive comments.
[18] The Editor thanks two anonymous reviewers for their assistance in
evaluating this paper.
References
Ammann, C. M., F. Joos, D. Schimel, B. L. Otto-Bliesner, and R. Tomas
(2007), Solar influence on climate during the past millennium: Results
from transient simulations with the NCAR Climate System Model, Proc.
Natl. Acad. Sci. U. S. A., 104(10), 3713–3718, doi:10.1073/pnas.
0605064103.
Bal, S., S. Schimanke, T. Spangehl, and U. Cubasch (2011), On the robustness of the solar cycle signal in the Pacific region, Geophys. Res. Lett.,
38, L14809, doi:10.1029/2011GL047964.
Bond, G., B. Kromer, J. Beer, R. Muscheler, M. N. Evans, W. Showers, S.
Hoffmann, R. Lotti-Bond, I. Hajdas, and G. Bonani (2001), Persistent
solar influence on North Atlantic climate during the Holocene, Science,
294, 2130–2136, doi:10.1126/science.1065680.
Braconnot, P., et al. (2007), Results of PMIP2 coupled simulations of the
Mid-Holocene and Last Glacial maximum—Part 1: Experiments and
large-scale features, Clim. Past, 3, 261–277, doi:10.5194/cp-3-261-2007.
Brown, D., R. Brownrigg, M. Haley, and W. Huang (2012), NCAR
Command Language (Version 6.0.0), Natl. Cent. for Atmos. Res., Univ.
Corp. for Atmos. Res., Boulder, Colo., doi:10.5065/D6WD3XH5.
Collins, W. D., et al. (2006), The Community Climate System Model Version 3 (CCSM3), J. Clim., 19, 2122–2143, doi:10.1175/JCLI3761.1.
Cubasch, U., R. Voss, G. C. Hegerl, J. Waszkewitz, and T. J. Crowley
(1997), Simulation of the influence of solar radiation variations on the
global climate with an ocean-atmosphere general circulation model,
Clim. Dyn., 13, 757–767, doi:10.1007/s003820050196.
Gray, L. J., et al. (2010), Solar influences on climate, Rev. Geophys., 48,
RG4001, doi:10.1029/2009RG000282.
Jiang, H., J. Eiriksson, M. Schulz, K.-L. Knudsen, and M.-S. Seidenkrantz
(2005), Evidence for solar forcing of sea-surface temperature on the
North Icelandic Shelf during the late Holocene, Geology, 33, 73–76,
doi:10.1130/G21130.1.
Knudsen, M. F., P. Riisager, B. H. Jacobsen, R. Muscheler, I. Snowball,
and M.-S. Seidenkrantz (2009), Taking the pulse of the Sun during the
Holocene by joint analysis of 14C and 10Be, Geophys. Res. Lett., 36,
L16701, doi:10.1029/2009GL039439.
L22602
Krivova, N. A., S. K. Solanki, and Y. C. Unruh (2011), Towards a longterm record of solar total and spectral irradiance, J. Atmos. Sol. Terr.
Phys., 73, 223–234, doi:10.1016/j.jastp.2009.11.013.
Lockwood, M. (2011), Shining a light on solar impacts, Nat. Clim. Change,
1, 98–99, doi:10.1038/nclimate1096.
Ma, L. H. (2009), Gleissberg cycle of solar activity over the last 7000 years,
New Astron., 14, 1–3, doi:10.1016/j.newast.2008.04.001.
Meehl, G. A., J. M. Arblaster, G. W. Branstator, and H. van Loon (2008),
A coupled air-sea response mechanism to solar forcing in the Pacific
region, J. Clim., 21, 2883–2897, doi:10.1175/2007JCLI1776.1.
Meehl, G. A., J. M. Arblaster, K. Matthes, F. Sassi, and H. van Loon
(2009), Amplifying the Pacific climate system response to a small 11-year
solar cycle forcing, Science, 325(5944), 1114–1118, doi:10.1126/science.
1172872.
Morley, A., M. Schulz, Y. Rosenthal, S. Mulitza, A. Paul, and C. Rühlemann
(2011), Solar modulation of North Atlantic central water formation at
multi-decadal timescales during the late Holocene, Earth Planet. Sci.
Lett., 308, 161–171, doi:10.1016/j.epsl.2011.05.043.
Park, W., and M. Latif (2012), Atlantic meridional overturning circulation
response to idealized external forcing, Clim. Dyn., 39, 1709–1726,
doi:10.1007/s00382-011-1212-0.
Prange, M. (2008), The low resolution CCSM2 revisited: New adjustments
and a present-day control run, Ocean Sci., 4, 151–181, doi:10.5194/os-4151-2008.
Shindell, D. T., G. A. Schmidt, R. L. Miller, and D. Rind (2001), Northern
Hemisphere winter climate response to greenhouse gas, ozone, solar, and
volcanic forcing, J. Geophys. Res., 106, 7193–7210, doi:10.1029/
2000JD900547.
Steinhilber, F., J. Beer, and C. Fröhlich (2009), Total solar irradiance
during the Holocene, Geophys. Res. Lett., 36, L19704, doi:10.1029/
2009GL040142.
Swingedouw, D., L. Terray, C. Cassou, A. Voldoire, D. Salas-Mélia, and
J. Servonnat (2011), Natural forcing of climate during the last millenium:
Fingerprint of solar variability. Low frequency solar forcing and NAO,
Clim. Dyn., 36, 1349–1364, doi:10.1007/s00382-010-0803-5.
Varma, V., M. Prange, F. Lamy, U. Merkel, and M. Schulz (2011), Solarforced shifts of the Southern Hemisphere westerlies during the Holocene,
Clim. Past, 7, 339–347, doi:10.5194/cp-7-339-2011.
Wang, Y., H. Cheng, R. L. Edwards, Y. He, X. Kong, Z. An, J. Wu, M. J.
Kelly, C. A. Dykoski, and X. Li (2005), The Holocene Asian monsoon: Links to solar changes and North Atlantic climate, Science,
308, 854–857, doi:10.1126/science.1106296.
Weber, S. L., T. J. Crowley, and G. van der Schrier (2004), Solar irradiance
forcing of centennial climate variability during the Holocene, Clim. Dyn.,
22, 539–553, doi:10.1007/s00382-004-0396-y.
Yeager, S. G., C. A. Shields, W. G. Large, and J. J. Hack (2006), The lowresolution CCSM3, J. Clim., 19, 2545–2566, doi:10.1175/JCLI3744.1.
6 of 6